Whereas AI seemed mostly experimental a few years ago, it is now a technology that corporate lawyers can no longer ignore. AI systems are getting better and better at processing texts, generating draft documents and supporting research. Moreover, it is expected that AI will continue to develop in the coming years and will be able to take over more and more tasks that are currently human work. Those who do not learn to work with it now risk missing the boat. It is always difficult to assess a technology that is still in full development. Nevertheless, we can now draw some tentative conclusions from our own experiences with AI.
First things first: AI is actually very useful. Reading about AI or taking a course can be valuable, but ultimately you only discover the real possibilities by experimenting with it yourself. That doesn't all have to happen in a business context at all. Indeed, it is precisely in personal life that it is easier to recognize applications. Plan a trip, make recipes based on the contents of your refrigerator, read menus in a foreign language, etc. Once you know and use the possibilities, application in work is a whole lot simpler and more logical.
Some approachable ways for lawyers to get started in their work:
Do market research on your client or counterparty. Ask an AI system to profile an industry or company. This will give you a quick first impression of their position, reputation or legal sensitivities. Of course, you have to verify the facts, but it can be a useful starting point.
Do research on your opposing party's market. We recently had to defend a client in a lawsuit alleging that his general terms and conditions were unusual and unreasonable. Using AI, within a few hours we were able to compile an overview of the terms and conditions used by competitors worldwide, in a variety of languages. This allowed us to demonstrate that the industry had similar terms and conditions around the world. Without AI, this would not have been possible, or at least not so quickly and cheaply.
Have AI identify inconsistencies in an argument. Take a pleading from the opposing party or a draft from yourself ask the AI where the reasoning is weak or inconsistent. It can reveal blind spots surprisingly quickly. This is a fairly safe application because it does not require in-depth legal knowledge. Knowledge of language, which AI systems are already very good at, is sufficient.
Let AI help you write publications. Talk to an AI about the article you want to write. Who is it for, what is the style, how long should it be, what themes do you want to name. Have AI write a first draft and work through it together until it is good enough to finish it yourself.
Let AI improve your texts. Legal texts are often long and formal. An AI can rewrite the same content in a more accessible style, which can be helpful towards clients, courts or internal stakeholders.
For those who want to go a little further, there are more complex applications:
Experiment with your own legal "agent." Put together a defined data set - for example, a set of judgments, policy documents or internal memos - and let the AI generate answers from that alone. In doing so, you mimic a junior digital employee who quickly provides the right passages. In this way, you learn how context constraints work and what the technology encounters in practice.
Train an AI in your own style. Especially if you use Copilot and it has access to your Outlook and documents, you can use AI to use a tone-of-voice that matches your own opinions, contracts and litigation documents. This requires some work, but is well worth the investment. It makes AI's input a lot more useful and recognizable, which can make your own work more efficient.
Self practice reveals where AI is already saving time and where there are still pitfalls. You learn to recognize the signals of a hallucination, or a convincing sounding but factually incorrect answer. You will experience how important good prompt formulation is, and you will discover in which situations AI especially speeds up and when it is better to keep the wheel in your own hands. Moreover, it gives lawyers a credible position towards (internal) clients and stakeholders: you can not only explain what AI can and cannot do, but also show how you use it in practice, and where the risks lie.
However, AI is not yet useful for strategic or truly substantive legal work. The quality of the output is too variable. Hallucinations are still too common and can often only be recognized if you have a really good grasp of the subject matter. For example, an AI that lists the most important lessons from a judgment usually gives partially correct answers, but can also introduce a completely new consideration or misreference existing passages. For a lawyer, this is particularly risky. Also, AI systems often just miss the point of a question. Therefore, it is no substitute for years of experience and the ability to weigh interests, strategy and nuance. The human factor remains crucial.
At the same time, it is clear that AI will only get better. Legal departments and law firms will therefore have to ask themselves where their added value lies. If the business or client can already go a long way in analyzing contracts or understanding regulations with an AI tool, what does the lawyer add? Incidentally, this issue is not new: the advent of the Internet also made legal knowledge more widely accessible. While this changed the work of the lawyer, it did not reduce the demand for legal services. On the contrary.
Therefore, our expectation is positive. AI can further improve the services of lawyers. Within the time or budget available, lawyers can do more meaningful work. They can emerge as the senior advisor who does the final check and adjusts at the beginning or the end, or who delivers better services in less time. We also see opportunities where tech-savvy lawyers use AI to optimize their services in such a way that the clients or business need to do it themselves to a much lesser extent. This returns control back to the lawyer, which also helps to manage risk. Herein does lie a clear assignment for lawyers: those who pick up the gauntlet now can make a difference later.
In the coming years, we will undoubtedly see further steps, not only due to technological advances, but also due to evolving legal frameworks. With the recent enactment of the AI Regulation, it is clear that legislators and regulators are stressing the importance of reliable and transparent AI. For the legal sector, this means a double challenge and opportunity: learning to work with AI, as well as ensuring that the deployment of AI itself meets the new obligations. There are interesting times ahead.
